Data mining? It’s not just a way for Amazon to better understand its customers’ buying patterns. A new research project at the U will apply computer techniques to poetry in order to visualize patterns and potentially give poetry scholars new insights about how poetic elements interact.
Katharine Coles, professor of English at the U and Utah Poet Laureate, and Min Chen, professor of scientific visualization at the University of Oxford, together won a 2-year grant from the Economic and Social Research Council’s “Digging into Data Challenge.” The grant is funded by eight international research organizations, and the Coles/Chen joint project was one of fourteen selected from an international pool to “investigate how computational techniques typically applied to the sciences can be applied to change the nature of humanities and social sciences research.”
Coles notes that the size of the grant—$125,000 for her part in the research—may pale in comparison to what a grantee in the hard sciences might expect to receive, but in humanities—and poetry in particular—it could be game-changing.
At a minimum, the interdisciplinary team aims to learn if computers can be utilized effectively for poetry scholarship. First, the poetry scholars will identify important elements of poems for the computer scientists to digitize and visualize. Once the data is visualized, poetry scholars will be able to use the visualizations to examine and compare such elements as meter and syntax as they operate within different poems.
Coles is a stranger neither to science nor to computing. Her writing often examines scientific themes, and she is married to an expert (Chris Johnson, director of the Scientific Computing Institute at the U) in computing. But using computer visualization to study poetry is new, although it may really just become another way—albeit more precise and granular—of studying language, according to Coles.
“These techniques will enable poetry scholars to look at works with a level of detail and objectivity that we’ve never had before,” says Coles. “What if we were able to analyze data to look deeply at poetry written at the end of one poetic movement, just as a new approach to poetry is about to arise,” she hypothesizes. “Could we use the insights gained from seeing what happens with the major writers to look for figures, perhaps not previously studied in any depth, who may be laying groundwork for evolving work?”
That is just one question poetry scholars may be interested in asking, should it prove that advanced visualization techniques are useful to humanities researchers. Because poetry is so complex, the poetry scholars will first break down relatively simple elements of selected poems—perhaps meter and syntax—for the computer experts to digitize, before moving on to more complex figures like metaphor. Coles notes that, beyond the complexity of individual poems, the body of poetry itself is vast. Eventually, once tools have been developed and tested, she imagines scholars will work to digitize and analyze poems going all the way back to the beginnings of English poetry, but for now they will begin at a time late enough that the language is similar to what we use today.
When asked if deconstructing poetry to such a mechanical a level would take away some of the “magic” that makes poetry, well, poetic, Coles replies with the artistic intention of a scientist: “It seems to me that the more we know, the more we learn about what we don’t yet know. The mystery just keeps getting deeper. In 50 years, how will the work and analysis that we do now change or enrich what is done then?”